MLX-Audio
Open-source Python library that runs text-to-speech, speech-to-text and music models locally on Apple Silicon Macs.
AI Audio & VoiceFreeverified Listed on Open source
Is MLX-Audio free?
Yes, MLX-Audio has a free plan. What you get: Open source (MIT license): free audio AI library you run on Apple Silicon Macs. To get it: install with pip or uv; then run a model from the command line or Python; then save the generated audio to a file. Verified on github.com.
Free plan at a glance
- Includes
- Open source (MIT license): free audio AI library you run on Apple Silicon Macs
Free and open source under the MIT license, run on your own Mac; the maker states no usage limit.
- Duration
- Free forever
How to get it free
Self-host- Install with pip or uv
- Run a model from the command line or Python
- Save the generated audio to a file
Good to know
- Apple Silicon Macs only
- ffmpeg needed for MP3, FLAC, OGG and Opus output
github.com, Checked Source
Overview
MLX-Audio is a free, MIT-licensed library built on Apple's MLX framework for text to speech, speech to text, speech-to-speech and music generation on M-series Macs. It supports models such as Kokoro, Qwen3-TTS and Whisper, with a web interface and an OpenAI-compatible API.
Key features
- Text to speech with models such as Kokoro, Qwen3-TTS, Chatterbox and Dia
- Speech to text with Whisper, Distil-Whisper and Qwen3-ASR
- Voice cloning and speed control on supported models
- Web interface and OpenAI-compatible REST API server
- Quantization to 3, 4, 6 or 8 bits
- Swift package for on-device speech on macOS and iOS
Best for
- People
- Mac developers running speech models locally, Developers who want an OpenAI-compatible speech API on their own machine, IOS and macOS developers adding on-device speech
Limitations
- Needs an Apple Silicon Mac (M1 to M4) and Python 3.10 or later
- ffmpeg is needed to save MP3, FLAC, OGG or Opus files; WAV works without it
- A code library: you run models from the command line or Python
Screenshot
FAQ
- Is MLX-Audio free?
- Yes. MLX-Audio has a free plan: Open source (MIT license): free audio AI library you run on Apple Silicon Macs. We last checked github.com's pricing page on 2026-11-12.
- What does the MLX-Audio free plan include?
- Open source (MIT license): free audio AI library you run on Apple Silicon Macs. Source: https://github.com/Blaizzy/mlx-audio (checked 2026-11-12).
- How do I get MLX-Audio for free?
- Steps: Install with pip or uv; Run a model from the command line or Python; Save the generated audio to a file. Start at https://github.com/Blaizzy/mlx-audio (checked 2026-10-24).
- Do I need an account to use MLX-Audio for free?
- Not for the free option: the free route is to self-host MLX-Audio (Install with pip or uv). Checked 2026-10-24.
- What's the catch with MLX-Audio's free plan?
- Apple Silicon Macs only. ffmpeg needed for MP3, FLAC, OGG and Opus output. Checked 2026-10-24.
- What are free alternatives to MLX-Audio?
- NaturalReader, TurboScribe and Notta all have free plans. We checked their free offers on 2026-11-12.
- What hardware does MLX-Audio need?
- The README lists an Apple Silicon Mac (M1, M2, M3 or M4), Python 3.10 or later and the MLX framework.Source: github.com/Blaizzy/mlx-audio
- Which models does it support?
- Text-to-speech models such as Kokoro, Qwen3-TTS, Chatterbox and Dia, speech-to-text models such as Whisper and Qwen3-ASR, and music generation with MiniMax Music 3.Source: github.com/Blaizzy/mlx-audio
- Does it have an API?
- Yes. It includes a web interface and an OpenAI-compatible API server for text to speech and speech to text.Source: github.com/Blaizzy/mlx-audio
Free alternatives
NaturalReader
Text-to-speech reader for documents and web pages.
Free forever, Unlimited listening with free voices; 20,000 characters a day with Lite voices
TurboScribe
Unlimited AI transcription powered by Whisper.
Free forever, 3 files per day up to 30 minutes each
Notta
AI transcription and meeting notes.
Free forever, Monthly transcription minutes
Fish Audio
Text-to-speech and voice cloning.
Free forever, Limited monthly generations
Typeless
AI voice keyboard that turns speech into cleaned-up text in any app, on Mac, Windows, Linux, iOS and Android.
Free forever, Limited words each week, standard accuracy
More free AI Audio & Voice tools
- Speechnotes
Free forever, Unlimited dictation in the online notepad and Chrome extension, with ads
- Abogen
Free forever, Open source (MIT license): free desktop and web app that runs offline on your computer
- whisper.cpp
Free forever, Open source (MIT): free C/C++ speech recognition you build and run yourself
- noScribe
Free forever, Free desktop app, transcription runs on your computer
- Spleeter
Free forever, Open source (MIT): free Python library that splits music into 2, 4 or 5 stems
- Podcastfy
Free forever, Open source (Apache-2.0 license): free to run; local LLM and Edge TTS need no keys
Sources
- Free offer
- Verified on github.com/Blaizzy/mlx-audio/blob/main/LICENSE
…is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation…
- Page content
Checked against
Added by our team. We check that a free option exists and record the date; signing up is between you and MLX-Audio, so read their privacy policy and terms first.
Is this your tool? Claim it · Promote it
